I always thought that would be a thing, glad to know it wasnt a wrong thought! Im not a machinist but I do watch ThisOldTony so that ofcourse makes me qualified xD
And then you need to calculate tolerances for tooling, and then tolerances for tooling for making tooling, and then tolerances for tooling for making tooling for making tooling...
"It's compounded by the fact that you will never be able to predict and test everything." This might be the wisest thing I have heard you say, Ian. I was an engineer for 45 years. I don't know how many times I tried to explain this to the man who just couldn't wait to start making money on something that seemed so simple to him.
@Fire Starter no idea what lasers have to do with engineering guns? As for computers, the most famous guns most people were designed before computers existed. But besides that, no simulation is perfect and nor can it predict some idiot hotloading his home-made ammo with dynamite or something else stupid. The GIGO principle applies at all times.
@Fire Starter you think a 3D printer is more precise than a CNC? Fine, it might have less vibrations, but at the very least the grain size is going to be a tolerance limit. Also, like Ian said, it comes down to cost.... printing individual guns is SLOW. For prototyping a 1-off, yeah it's faster than spinning up an entire mold/press/ jig, but once it's going it never gets faster.
I keep having to explain to the 'young' guys that Murphy Rules With An Iron Fist. You plan for everything you can think of, and maybe you succeed, but then it is the thing that hits you sideways that you never thought of that derails you. Example: Years ago I did a live broadcast event in a country that, shall we say, wasn't exactly known for freedom. Everything was lined up with umpteen levels of redundancy. Event day came and the government official who had helped us with "things" was nowhere to be found. He defected to the rival country next-door. Needless to say, things didn't go well after that. You learn to plan for the unexpected - always have a disaster plan and always expect that something will go wrong. Most of the time you're fine, but the Universe seems to like to spring things on you at the worst possible time and place. ;-)
@Fire Starter 3DP metal often isn't very durable as it's prone to microbubbles that propagate fractures. It's the most loose and awful way to assemble a metal part, which is why we just don't use it to make anything that forms a pressure vessel, (like the receiver of a gun does while firing). High end 3DP metal, such as used by SpaceX for the turbopump propellers in the Merlin rocket engine, are extortionate. It's used there because the primary demands are light weight, high heat tolerance, and very, very precise measurement tolerance. A mas-produced gun using 3DP parts would be far more expensive than the competition, and probably still prone to failure due to the lack of structure in its metal components. If you want to learn more about this, Real Enginerring did a great video with microscopic footage of the laser sintering process showing the trapped air bubbles. It's like a part made of the sloppiest welding known to man. Fine for making anything domestic or aesthetic, useless for most heavy applications.
To me the fact that those large caliber break actions dont explode at the breech end every time you shoot them is simply incredible. Given that the barrel and breech are virtually flush with each other with no recessing of the cartridge, every shot is a mechanical miracle.
As someone who programs/sets up CNC mills and lathes, I can absolutely appreciate this video. Many ppl underestimate how much truly goes into manufacturing precision parts.
Yea, I work in tool maintenance for a company specialiseing in stamping and deep drawing. All those issues are realy familiar with me. Especialy the whole "figure out the cause of some weird, random malfunction" part.... :-D
I worked a factory job in a tier 2 auto parts manufacturing/painting factory. A large paint line, cream dip oven bake type. We painted rear axel housings, you know the differential bell housing with axel tubes and many attachment brackets. One station, pressed in the bolts for the diff cover and then the axel housing went on a stand where the housing can easily be rotated. In one hand you held 2-3 no go/go gauges and an orange dead blow nonmarring hammer in the other. You adjusted these brackets to easily accept the gauge, each gauge was rectangular able to gauge two brackets each. Once all brackets were adjusted and welds examined specific points were marked with paint markers. These type of factories have whole Quality Control departments, whose sole job is to make a round daily taking parts off stations and clamping them in a QC fixture to ensure tolerances were in spec. These people had authority to stop any production immediately and technicians would identify the issue and make adjustments/repairs to fix whatever issue was identified. All parts containers are labelled with station, person and date/time information in code form. Parts that could be out of spec were sorted and any that were separated and put in waste/recycle bins. As design and manufacturing equipment gets better and better, accuracy is improved in parts production and overall design to production times have to deal with how much money is potentially being made contrasted with the costs of r and d plus production. Things could be very rapidly be mass produced accurately if cost were of no concern. People want return on investments not simply producing the highest quality masterpiece priced in range everyone could afford. Hopefully we’ll get there one day though.
Heard fingers , long hair & eyes horror stories from our shop class teachers in high school, getting fingers /hand lopped off or hair /scalping in a drill press. Eyeballs popped etc. Safety First☝🏻
The tooling and tolerances is the main problem in many stories. Then there might be the engineer who has forgotten the difference between US and British inches, let alone centimetres.
The US and British inch, as well as the Canadian inch, became the same decades ago. Certainly a thing a good while ago, but not at present day. Assuming modern metrology equipment of course.
That's why you check tolerances every X cuts or X amount of guns. The problem comes from how much is too much or too little and how often do you check. That's one of those never ending problems unless you are working with indestructible tools.
The difference was more critical in the 19th century - the .303 rifle was the result in the difference between the Enfield inch and the British standard inch. Enfield was using a standard 1% longer than the British standard. I am aware of 40 different inch standards in use in Europe prior to the metric system being adopted. The American inch has been a metric derived unit since 1959 and the old definition is due to be finally legally phased out in 2022. (Legal American inches prior to 1959 were slightly longer and British inches slightly shorter than the current standard but machinists had been using the modern standard definition since the 1930’s).
interchangeable parts, production line, reliability, efficiency, cost-effect those are very difficult problems, people are tend to ignore those problems since we're already in a mass production-consumption era, but the problems are still there and requires great deal of time and money investment with social scale corporation
Well they had to put through every weapon to town on Mars dude God knows how long it took the lasrifle to work right still waiting on my mastercraft hellgun bro I hope when it does come out someone uses it to serve the emperor right tear against chaos and xenos.
@@HBK-6S The joke is that modern civilisation has lost the specs the "STC" equiv of some ancient but venerated peice of tech, making modern gun nuts somewhat similar to the Admech in a way, not that the Admech in 40k literally wants the STG desighn.
I've done production machining. So I know what you are saying about getting all those tolerances right. There are other factors that come into play during the production run tho. Tool wear is the most common. Then there's tool spring. The tendency for the cutter to bend away when pressure is applied to it. Then there is chip control. How to keep chips from building up and interfering with the cutter. Then there is tool breakage. In a CNC machine this can be disastrous if a drill breaks, and then a tap comes along and tries to thread that hole! Then there is the engineer who draws a perfect part only to find that what's on paper doesnt necessarily work in real life. Tight tolerances may be needed, but after shooting many rounds in a row the metal expands and those tolerances will come together and jam up the works. And when you have everything tuned and running right, some plant manager will come along and try to change something. Starting the whole process of tuning the machine all over again. Spending a week breaking tooling, producing scrap, and wearing out tools faster. And all just to save 1 and a half seconds of machine time.
Tool wear is also a major challenge. My dad was a Master Tool and Die maker part of his early work was making and resharpening cutting tools for the production line along with six die sets to crush couplings tight on hydraulic lines used on big earthmovers. Tolerances on some hydraulic parts can be in 1/10,000 of an inch plus minus 1/100,000 range.
This is why Brandon Herrera gets so frustrated with people constantly asking when the AK-50 will be ready. Designing a gun is hard enough. Designing one around the pressure generated by .50 BMG is on another level. You can't just scale it up; you can't scale the molecules of the metal up. More massive parts with higher stress on the same size molecules can do some janky stuff, and janky combined with explosives can be a really bad time.
As a retired automotive product designer, I can relate to everything in this video. In my 40 years of design work I started with hand drawings and at the end was doing solid modeling on computers. Most of the work I was doing wouldn't get into final production for 4 or 5 years.
Ian - I’ve learned so much from your channel, but THIS is the best presentation you’ve done yet. Tolerance stacking can generate really elusive gremlins in any complex machine. I really like KelTec’s innovative product ideas but perhaps they should hire you for process review and QC. When I bought a PF-9 the salesman told me “every once in a while they spit out a lemon. Just bring it back.” My initial incarnation literally couldn’t get through a single magazine without jamming. We sent it back. 10 days later I got the original frame back with virtually everything else replaced. It has worked flawlessly ever since, but someone had to take the time to find replacement parts that got an A in “plays well together.” I would love to see you do an expose on the “you need a 500 round break-in period” BS from some manufacturers. I got that from Kimber on a 1911 with an extractor that could clock roughly 20 degrees coupled with a firing pin stop that literally flopped around in the slots at the back of the slide. Assuming you could get through 500 rounds without losing your mind, it would have only made the problems worse due to excessive wear on the extractor claw. 50-100 rounds for an extremely tight gun produced for gun range accuracy, OK. Breaking in a rifle barrel is a real thing IMHO; however, that’s completely different from a gun that jams right out of the box. Frankly, I think the 500 round garbage is to deflect the many people who buy a gun, maybe take to the range once or twice, and then never fire it again. (And if I’m wrong in my assumptions and assertions, please correct me!)
As a machinist, and hobbyist gun maker and inventor, I have to give Ian a lot of credit for taking a very precise science and, in this case, ‘dumbing’ it down for the masses. He eloquently and concisely gets the point across while adding just enough fact and history to keep it entertaining. That’s the whole reason why I subscribe. That, and to give me ideas on what to build or mimic in my next pewtoy.
This is a significant factor in the failure of Hudson and the H9. They released a product where issues cropped up and they didn't have the capital to fix them while also paying off their debts. In fact they were far enough behind that some of their suppliers sued them for failure to pay after having already delivered parts. The lesson to be learned is that whatever you are making, assume you will spend multiples of what you planned getting it done before you turn a profit. This is true outside firearms, as well. When making the original Star Wars film in the 1970s the studio was willing to put up $8 million to make it. George Lucas spent almost $12 million and still wasn't happy with it, in the end. Lucky for them it went on to be the most popular thing ever and they made their money back (and then some). There are also movies that look like successes when you compare production budget with ticket sales but then you look up the marketing budget. Several of the Fast&Furious movies spent more on marketing than on making the movie and those were NOT budget films.
Excellent video! As an engineer, I would be interested in a video about what new technologies can help to shorten the times from concept to production (CAD, CAM, 3D printing, etc.) and what example there is that new technologies have achieved. . Some examples of the WWSD process would also be good. Greetings from Argentina
I've been involved in productionizing mechanical products for 40yrs so I agree with all you've said. I would like to see you explain something like gun barrel design & manufacture. The history of guns tracks the metallurgy of barrels & the ability to do accurate gun-drilling. One point in your explanation was on tolerances, all good but the build up of tolerances is important in a mechanism. All components can be in tolerance, but if all at bottom end the assembly can be off.
As a CNC machinist, this is why I love your channel and love guns. I feel like any person in manufacturing has got to appreciate a well designed and built gun or automobile.
I think I enjoyed this video more than any other! In an internet of oversimplification, here's a breath of fresh air. I spent a lot of my career in production, and yep, it's hard. Just making the production tools run is hard, figuring out what they need to be making ~exactly~ is even harder. Excellent video, excellent examples! 8^)
There's a gun in my collection that I think illustrates this problem quite well. Back in the '90s, a company down in Texas made a Luger clone in stainless steel. Unfortunately, as many of you may know but they apparently didn't, stainless steel isn't a drop-in replacement for regular steel in mechanisms with moving parts and bearing surfaces. So even if they had the entire Luger TDP (and they probably didn't), if all they did was make the parts to the original specs, it was never going to function properly... as indeed it does not, and never has, even after it went back to the factory for modifications shortly after I bought it.
Ian there was only talking about dimensional tolerance. That leaves out material composition heat treat tolerance coatings tolerance... You can nail all the dimensions and have a part be a little too riddle or a little too ductile and it won't do its job. Or maybe it does his job just fine but it beats up the part next to it. A good example too is minor tweaks making big differences. There's a series on SOTAR where he shows differences in metal surface treatments. And similarly on small armed solutions. An AR bolt carrier that is dimensionally identical but nitrided instead of phosphatid is a little bit more brittle in one impact point so they have to change the dimensions very slightly to get the same wear properties. And in a lot of cases when they take the same thing and make it in nickel boron coated, then that can mess up the screw threads for the gas key and cause leaks. So they have to add a manufacturing step of chasing the threads. Many manufacturers don't do that which makes nickel boron carriers potentially less reliable even though they on paper are better. If somebody does the extra step of chasing the threads and making sure the gas key is well sealed then they probably are better.
This video is excellent because it applies to far more than just guns. Aircraft and warships are two other common examples where we see these happen. Aircraft generally have prototype and pre-production versions, with the early aircraft partially handmade and partially using the new tooling. The F-35 has been criticized as the early aircraft will only be used for testing and training, but spare parts will not fit as precisely in later aircraft and there have been tweaks since the early low rate initial production aircraft. Ships rarely have a prototype built before the rest of the class, but lead ships in general have far more defects than the subsequent ships. One noted example of a prototype ship are the first two Littoral Combat Ships, built using R&D funds and with significant differences from later ships. Notably, Independence had a problem with galvanic corrosion in the water jets, corrected on the next ship, but to this day you’ll hear about how the entire class cannot last in salt water because it’s made of aluminum. The recent confirmation that the first two ships will be decommissioned has many attacking the ships as garbage, but the plan for the last several years was the first four were supposed to be an R&D division because of their differences from later ships. Regarding tolerances, one key example I know of is the Type XXI U-boat. This is widely regarded as one of the most important submarine classes ever built with almost every post-WWII submarine incorporated lessons from the boats, but the subs were built in sections by different manufacturers and shipped to the yard for final assembly. The first Blohm & Voss boat was so bad it became a dedicated training ship at the yard, and working out the proper manufacturing tolerances prevented hundreds of submarines from being completed on schedule (a bigger problem than even Allied bombing) and kept most of the 118 completed from going on patrol. The US, by contrast, built several ships by modular construction during the war, with far fewer issues. However, certain shipyards built better quality ships than others. Cramp built submarines often had to be completed at a different shipyard, and a certain production block of Electric Boat submarines (centered in the 340s) were preferred for the best postwar conversion programs. Manitowoc could not transition from the Gato to Balao class on schedule because they lacked production drawings, though they generally built good boats. For destroyer escorts, the Dravo, Wilmington built ships were preferred for foreign sale during WWII, cancelled unusually early during the phased slowdown, and sold off extremely quickly postwar even compared to other Cannon/DET ships. For material selection, Little Ship, Big War makes an interesting point on US DE construction. To speed up steel production (which was a bottleneck during the war), the new steel plates were not given the final Bath to remove the mill scale of impurities on the surface of the steel. This steel rusted far more readily than normal, passing the burden onto the DE sailors to scrape and paint. For Abercrombie, Pete Kish waged war on mill scale during a stay at Manus, and after 25 days the entire stern of the ship had been scraped to the bare metal before being primed and painted with “several” coats of primer and two coats of deck blue.
Great video! Most of this applies to just about any mass--produced commercial product, too. One thing you touched on is the relationship between product development and process development; sometimes the product design drives process development, and sometimes it's the other way around; you start with the process and develop the product around that. Another concept you touched on is "good." We consumers tend to judge the quality of a gun based on things like how well it shoots, etc. But that is meaningless if it doesn't make money. A manufacturer has to look at profitability first - if you can't sell it at a profit, it doesn't matter how "good" a particular gun design is, it won't be made on a large scale. We tend to think we can "figure something out" just by looking at the end result, but that's hubris. Sometimes the reason for a particular design element is so far removed from the final product it's hard to make the connection at first glance. And most people have a hard time understanding what it takes to go from hand-crafting one item to having a factory that spits them suckers out by the pallet-load.
I do Mechanical Design for a living. This is a fantastic way of explaining this. I try to explain it to people all the time. There's SO many things that the average consumer doesn't understand. Let me tell you, as a designer, determining the tolerances and fits is really difficult. Even with CAD and simulators, we don't always know whats going to happen. Great video Ian! Also, people think CNC processes are perfect (Laser cutting, 3d printing, 5 axis milling, etc); but, depending on what you're making, they're not. Most machines get you close, and require secondary operations to complete.
While this video lacks "cool gun", the content is fantastic in regards to how the cool gun is made. So many people overlook the tool-room to production transition and the creation of jigs, gauges, etc. that are so vital to a successful line. The other thing often missed is the incredible expense associated with buying and installing all the machines to do the single operation functions. It takes big money to pull it off. And that's why so often we hear of the inventor and the backer in our history of guns.
Thanks for making this, transferring from prototype to mass production at high yield is an important under-appreciated field of engineering. Much love for explaining engineering concepts such as tolerencing and step yields.
I worked at a construction mobile crane manufacturing company in the US and I can tell you that in 2007 we probably produced the best quality product that we ever could and it was actually amazing. We are running 3 shifts 23/7 and every unit was already sold. All of our vendors were producing only for us our rejection rate was reduced to almost nothing no personal movement in production, no pressure for cost reduction at all no pressure for new design development all of engineering was focused on production throughput not on cost. It was amazing everyone was happy if a little tired. Everything was running better than I thought even possible. It was really a sight to behold. And then 2008 and every thing went to hell even though production was cut to less than a quarter of the year prior our quality was terrible by the end of the year.
The official issue of the M4 Block 1 vs the M4 Block 2 vs the M16A2 is an interesting case in "reinventing the steel" to get it right for mechanical variances, recoil impulse, dynamics in the DI, etc.
Dimensions and tolerances are great, but there are lots of other things. Metallurgy comes to mind - heat treating, bluing, residual stresses, voids, getting the forging right, getting lubricant that matches the needs of the weapon. And then when you get that all straightened out, some sailor drops your pistol, the safety mechanism fails, another sailor is killed, and production of that sweet old Smith and Wesson trigger mechanism gets stopped.
As a machinist, I can attest that this is absolutely correct! We make aircraft parts and whenever we get new prototype parts, it always, ALWAYS takes time to figure out what tools we need to use and which tools are best. Even with our cnc machines, that takes time and rejected parts to figure out and problem solve!
One of the weird "little things" that typically happens with military firearms manufacture: fastener tolerances. In general, pressure tolerances due to tension forces (i.e., how much a fastener is pushing / pulling materials, causing some distortion) are tricky to figure out, and non-obvious; a firearm w/ perfectly-cut, bent, welded, etc. parts may be a huge mess if the fasteners are causing dimensional drift after final assembly (or, in some cases, during field-stripping). This is one of the reasons why really famously-reliable stamped firearms tend to have tight tolerances on a few small areas, but surprising forgiveness elsewhere (I doubt if there are two Stens anywhere that match within a thousandth on all major dimensions). Also, one of the fascinating bits about the AEK-971 system and the AK's based on BARS systems is just how long it's taken to get the BARS pulleys exactly right. This is probably a textbook example of teething issues in a new concept- even though the basic idea's been around for 50 years! And yeah, the AK-50 is still not even an alpha-stage firearm. But I honestly think Brandon will eventually get it "working", in the sense that it will cycle properly. I think he's on the wrong track with the magazine, but hey, time will tell.
Great video! I think there's almost no knowledge in general public of how complicated it is to mass produce complicated stuff. How every tolerance, every inefficiency can affect your manufacturing process. Really appreciated that someone is talking about it.
This is fantastic information. I occasionally tinker with my own gun designs, and even making one custom build work properly is difficult enough. I can't even imagine how difficult it would be to make them on a large scale.
One of the main topics for a term when I was at college. Reduced to a minimum, it goes like this: Choice of fit. Interference, transition or sliding? Then Allowance. How much space (Positive or negative) is needed between parts to work as you want them to? Then Tolerance. How far can each individual part be from the correct size and still work with the other parts (Maintaining the allowance)? Then Gauging and sampling. Work out how you are going to measure each part, how you are going to control the accuracy of your measuring equipment, and what sampling regime you are going to use. Then. Multiply by the number of parts. Then Cry.
Military small arms development may have also reached a point where there's no room for practical improvement to the technology. That may be why so much attention has been given to ammo development recently , which also has limitations. That's also why the Army can't seem to find a replacement for the M4 rifle that's improved enough to justify the cost.
Small, incremental improvements are still possible and being made: In metallurgy, in machining, in plastics, but nothing revolutionary, and it takes more and more investment to get smaller and smaller progress. Something like ICE cars in recent decades - companies have been shaving the tiniest improvements in engine efficiency, aerodynamics, production costs, and all the cars start to resemble each other, regardless of the manufacturers.
Hatchers Book of the Garand. Good reference on the m1 in general, and early US automatic rifle development too. Basically the follower would not apply correct support if the clip was in with the top round on a certain side. They fixed it by changing the follower geometry and giving it a small amount of rotational movement. At least as I recall, I don't have the book on hand. I do know it took them years to figure out, and this was after it had been adopted and was in production.
Sig employee: “hey boss should we figure out the tolerance zones before we produce and ship out all these guns?” Sig supervisor: “Tolerance zones? We just measure all the guns our beta-tes....I mean...our customers send back and sort it out from there”.
Just had a couple of webbinars dedicated to simulation in CAD. It allows to "look into" working machinery without building mock-ups and prototypes "in flesh", yet it isn't cheap and one got to be a pro to be able to model it. On the bright side, there are more and more cloud services for that, so you don't need a super-powerful desktop (which then also needs upgrades) of your own and expensive software, you can now rent time and power of cloud service for simulations, which still isn't cheap, but got to be _cheaper._
"But with modern manufacturing/materials/cnc machining/etc, gun that I want should be made cheaply and quickly" is my favorite thing that pops up on forums Yes, that 20th century gun or early aughts prototype that was originally designed for a long extinct or proprietary cartridge that also had a pretty specific pressure limit could be made yesterday in current cartridge offerings with enough made to not cost like 2k+ per
That was a really good explanation of the complications with development. CNC machines have to be programmed, and before that parts have to be engineered. I have seen firsthand business go under by producing over engineered parts. Sure it works, but the firm isn’t making enough. Also, there are so many things that can go wrong during development, and private firms are spending money, not making money, and do not have unlimited investment from government etc.
depending on the time period. there where semi automatic machine tools. starting around the 19th century, which could make numerous cuts in one machine, but still controlled by a man. Examples are, turret lathes, screw machines and semi automatic milling machines.
Interesting fact about car manufacturers. In some cases the same model goes for a few years, all ford Falcons in XD-XF range look very angular. Then a big change to the rounded Falcons of early ninites. While Nissan changed each year. One factory ran presses till they wore out, others reconditioned the press each year. So Falcons show the wear, XD-XF parts interchangeable but XR series not so much. Skylines changed ever so slightly each year and parts just don't quite fit. I imagine with pressed gun parts it's even more pronounced and less forgiving. Presses and dies wear out, every part made slightly changes the tool.
As a mechanical engineer and firearms designer, who has been working on a project for the better part of 10 years, I can tell you all that what Ian is saying is 100% correct. Even in my professional capacity at the company I work for, making automation equipment, we rarely get it right the first time. That isn't to say that we can't make it work, but that going from CAD to physical parts almost always turns up and unforeseen issue. Couple that with designing something that is meant to contain a controlled explosion and put a projectile on a consistently repeatable flight path, and you have your work cut out for you. From predicting failure points, to determining wear parts, all the way to assembly/disassembly and cleanability, guns are complex machines, meant to control extreme forces in a compact package.
Apparently not because they're all in the comments shouting about how this is iAn DeFeNdInG bRaNdOn, rather than actually recognizing that between SMG, HMG, Hudson Mfg. and Jim Sullivan interviews, Ian has probably had this video coming for a long while.
This video, in particular the parts about tolerances, offers a good explanation of why reverse engineering is not as easy as people think, and why knock-off parts often don't fit very well. Do you have a background in manufacturing industry, Ian?
M1 7th round stoppage was caused by the forward left guide rail being nipped off in production, leading to occasionally a round being loaded from the right hand side to feed incorrectly into the chamber.
Seventh round stoppage. Not in development but in early production. When boring the receiver for the barrel they bored too far. They cut off the upper part of the cartridge guige. This was for production below 40.000. See Hatchers Book of the Garand. They welded up the receivers affectec and loaded the clips so the seventh round was on the opposite side. Ps the cut /mistake was on the right side. Forgive spelling!
The L85 is a great example of how a problem here can result in a sympton waaaaaay over there. My dad found a production problem with tape recorders that was actually an issue with the lack of a toaster oven in the break room. How did you get from one to the other? Easy, just follow the tape head line through two shifts on a Tuesday or Thursday. No problem at all...
You can never predict everything. The most common fallacy in mechanical machinery is that people assume there is a direct cause and effect chain. A happens, causes B, which causes C, which causes D - so knowing A, you'll know that D will happen. Newtons laws taught in University confirm this theory, but it's often forgotten to mention what assumption precedes any application of the analytical solution of these laws - the assumption of perfect information. In reality, A might cause B, which causes C and then D. But A might as well Cause E, which causes F, and F prevents C from happening - but you never knew E and F were actually a possibility, as they don't even appear in your model. A great visualisation of this principle is the double pendulum. For every initial condition, there is a correct calculation of it's path if it is dropped, that can be calculated precisely. But in reality, the initial conditions vary ever so slightly, leading to a chaotic pattern of movement. Those are two (presumed) solid bodies in a mechanical relationship. A gun combines dozens. A lot can and will go wrong.
Thanks God you said that gun designing is a hard job As a gun Smith I really know how hard is it . It's almost my third year of designing A new pistol for accurate long range shooting
2:10 - "And then you take one of each part and put together a gun and hopefully it works" This is literally what Remington did with the R51. And, yes, I do mean literally and not as a joke. When Remington had to shut down R51 production (...the first time), they actually stated the reasoning for the issues with the R51 were because: "performance problems resulted from complications during our transition from prototype to mass production." So even a multi-million dollar company that *_should_* know better actually ended up killing off their company because they botched one of the most anticipated handgun releases at the time and never recovered from it. I know, myself, I thought the R51 was looking pretty slick and I was eager to see them come out. Thankfully I waited until other people bought it and ran into the issues so I didn't have to. And it's a shame, it could've been a really good handgun and gotten Remington a lot of sales that they desperately needed.
Maybe a good (part 2) of lecture would be to show a production line of the milling machines making 1 tiny part. Showing how the part is measured , gauged, accepted or rejected. I asked an mechanical engineer why a Sig 9mm was much more expensive than a Taurus 9 mm. He said that the steel used in the Sig was much higher quality than metal Taurus. This steel is harder to machine.
After watching this clip, it also ties with my major search of adopting my favorite musician. From 2009 to 2019, I went from Lady Gaga, then Britney Spears, and finally to Jessica Simpson because I kept changing my mind of what’s really my go-to musician. The major malfunction being is that after my sleep times, I lose interest over time and it happens to lots of my candidates. Then when Jessica Simpson comes into my interest, she finally solved the issue of my interest loss, thanks to the vast amounts of my life force multipliers, so I formally adopted her as my favorite musician and she will continue to serve my musical fandom as whole, just like the M16 and M2 Browning will continue serve our GI’s today.
It is interesting today. I was listening to a drachinifel video talking about the zero, which was a first rate, was designed by the Japanese. It was not a copy of another design. However, they were heavily dependent on US machine tools for mass production.
7:40 Also what seems missed. If kept within reason of the margins instead of being "perfect ie 1%" can also give tolerances that can improve maintenance requirements. Perfect meaning tighter tolerances that can easily foul during operation w/o proper upkeep. Looser ie 10% being more forgiving.
The powder issue with 5.56 rounds and the AR-15 reminds me of the Desert Tech MDR, since they had never tested with cheaper surplus .308 ammo, they didn't know those rounds would have issues when ya'll tested the gun with them.
GD&T is a great tool for getting things accurate across parts and keep your tolerances only tight where needed. For all other potential gun designers out there
If any gun manufacturer wants to leap ahead of everyone else they have to change the paradigm. Design things in a completely different way than before. If not they are stuck with a development curve that has already reached its upper limit. So any improvements are going to be small.
Knights Armament developed their light machine guns very fast. They say most of the design was done on a computer. But, it took years of knowledge and experience before those programs were even written.
I feel like a lot of this notion of "why don't they just make X" has been warped by the parts kit culture of the US market. Everyone knows someone who's done a parts kit--it's not super hard usually--why can't a company just do that from scratch? Difference is, some dude building a Kalash in his garage has theoretically unlimited time and an incredibly high budget, has the specs for a known product already at hand, and can debug his individual weapon over time. Making a new design involves the exact opposite of all of that
Random info regarding logistics of why you cant get ammo right now or why it is so expensive and also in tangent to Ian's point about how long it takes guns to get produced etc. concerning ammo and gun shortages and other shortages you might be experiencing in your area. In logistics or shipping goods via rail, air, and over the road (Trucks).... only the government and a handful of "carriers" (shipping/trucking companies) can broker/move these loads. These meaning- Drugs (prescriptions and also marijuana where legal), money (coins or bills), and firearms unless deactivated or going to a military base and also deactivated/unarmed as its usually empty shells for bombs and stuff that they will arm there. So guns, money, and drugs are the big 3 and then theres lots of restrictions on live animals, eggs, hazardous mats. and lots of other goods you might notice absent from their normal prices or shelves. Also right now there is so much freight drivers call the price, a load of boxes (15k lbs) from CA to VA is 8k and about 4-5 day transit while VA to CA with same details is 3k. The market is crazy, and dictated by goods, location, and driver availability. Drivers hate NY so goods there are always more expensive and harder to get trucks up there, also there is more freight in CA and in TX than anywhere so loads leaving CA or TX are going to be very expensive especially compared to inbound freight. The truck market is just like the stock market its always changing, governs a lot of things, and if you knew how deep it went youd understand why those strawberries are on sale at your super market (tip - dont buy them(well maybe not the strawberries but be weary of meats and lettuce)...they probably came in under temperature or something and the store accepted the shipment but filed an insurance claim on the load but they still just sell it off at a discount or toss it out if its too bad to sell.) Trucks and logistics govern this country and if a ship getting stuck in a canal didnt help anyone realize how fragile our current logistics system is...were in for trouble. Lets hope for better infrastructure but in the mean time get educated on your local store or restaurant's deliveries and how they can impact your life so maybe how to better be self sustaining. So next time theres a rush for TP or 9 mm, youre set.
you skipped on temperature expansion changes as it heats up from use. or cold weather, all these working gaps are effected from temperature that can jam a gun.
I can really see your engineering background coming through. I am super interested in this kind of technical knowledge and looks like a lot of other people are too. You should really make more videos like this!!
another thing about CNCs is if you move the machine from one location to another it will be off. i had a dept manager that couldn't understand. a guy went to where the machine was built and set it up and ran a certain number of parts to prove it could run the spec. after moving it over 300 miles to the factory where we worked . it was off and i had to reset it up. i expected it . the manager did not. getting gauges made takes time too and $$. good vid
Great vid Ian. Unless it's urgent design takes along time and a lot of unknown problems can occur and have to be solved before even simple products are ready for the open market.
Toleracing engineers are focused on dealing with cumulative tolerances. Having having tables of standard tolerances helps a great deal. In design the rule of thumb is to keep tolerance as lose as possible where possible. As well as the tooling/jigs and fixtures, there is also a need for the attribute gauging and inspection requirements.
This is a marvelous example of tha scale. When you have to make one gun work, it is easy, but, when you have to manufacture 10 mln. of guns, with scrape rate on the leve for example 5% it is hard as fuck. That's how the mass production works. It take's time, effort and money.
Making the machine that makes the machines that make the mass produced product is hard. Now make that product a gun that needs ammunition that can also vary in tolerance. I used to work in a factory running gun drills, punches, Acrilocs, Warner & Swazey chucker- drills and that was making brakes cylinders for military vehicles. I'm guessing that is similar and just as important asmanufacturing guns as far as tolerances are concerned. Lots of machines lots of testing lots of rejected parts. Later I assembled those same parts I made. I was glad the engineers figured out the tolerances then. The technical package is extremely important.
You could write an entire book (and people have) on minimum and maximum material conditions and how those interact with wear on interlocking moving parts over time.
